research article mining for candidate genes...

Research ArticleMining for Candidate Genes Related to Pancreatic Cancer UsingProtein-Protein Interactions and a Shortest Path Approach

Fei Yuan1 Yu-Hang Zhang1 Sibao Wan2 ShaoPeng Wang2 and Xiang-Yin Kong1

1 Institute of Health Sciences Shanghai Institutes for Biological Sciences (SIBS) Chinese Academy of Sciences (CAS) ampShanghai Jiao Tong University School of Medicine (SJTUSM) Shanghai 200031 China2College of Life Sciences Shanghai University Shanghai 200444 China

Correspondence should be addressed to Xiang-Yin Kong xykongsibsaccn

Received 6 September 2015 Accepted 15 October 2015

Academic Editor Jialiang Yang

Copyright copy 2015 Fei Yuan et alThis is an open access article distributed under the Creative Commons Attribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Pancreatic cancer (PC) is a highly malignant tumor derived from pancreas tissue and is one of the leading causes of death fromcancer Its molecular mechanism has been partially revealed by validating its oncogenes and tumor suppressor genes however theavailable data remain insufficient for medical workers to design effective treatments Large-scale identification of PC-related genescan promote studies on PC In this study we propose a computational method for mining new candidate PC-related genes A largenetwork was constructed using protein-protein interaction information and a shortest path approach was applied to mine newcandidate genes based on validated PC-related genes In addition a permutation test was adopted to further select key candidategenes Finally for all discovered candidate genes the likelihood that the genes are novel PC-related genes is discussed based ontheir currently known functions

1 Introduction

The pancreas is a significant versatile organ of the digestivesystemand endocrine system it is significant because it assistsin digestion and maintains hormonal balance via digestiveenzymes and specific hormones respectively Pancreaticcancer (PC) originates from the pancreas has a high rate ofmetastasis and is a highly malignant tumor derived frompancreas tissue Among various types of PC pancreaticductal adenocarcinoma (PDAC) accounts for more than 90of all pancreatic tumors PDAC is a malignancy with apoor prognosis which is demonstrated through its one-yearsurvival rate of approximately 18 for all stages of the disease[1] In the Western world pancreatic cancer is one of thetop killers for human beings [2] In 2012 alone it resultedin 33000 deaths all over the world In the Western worldpancreatic cancer is the fourth leading cause of death fromcancer with a poor prognosis (5-year survival in less than 5of cases according to most reports) Such a high fatality rateis attributed to the low rate of diagnosis at early age Only aminority of patients can receive proper treatment with a 5-year survival rate up to 22 Therefore it is significant and

crucial to study this severe disease Similar to most types oftumors PC is induced by both environmental and hereditaryelements Extrinsic factors such as age gender race cigarettesmoking and obesity are all factors that may contribute totumor initiation [3ndash5] Further certain chronic pancreas-associated diseases such as diabetes mellitus and chronicpancreatitis are also related to PC [6 7]

Over the last decade the genetic background for PCespecially pancreatic adenocarcinoma (PAC) that comprisesmost cases has been revealed through validating a list ofoncogenes and tumor suppressor genes for PC Based onhereditary features mutations in pancreatic adenocarcinomacan be divided into two clusters defined as common somaticmutations and germline mutations A somatic mutationKRAS is regarded as the earliest and key mutation in non-familial PAC initiation and it aids in maintaining invasionstatus and tumor progression [8] In addition to tumordevelopment more mutated genes contribute to malignantphenotypes The tumor suppressor genes p16INK4A aresignificant somatic mutations and are downregulated inpancreatic adenocarcinoma [9 10] In many types of tumorsexcessive activation of the TGF-120573 pathway is a mechanism of

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 623121 12 pageshttpdxdoiorg1011552015623121

2 BioMed Research International

tumor progression and invasion [11] Another pair of tumorsuppressor genes SMADDPC4 is involved in PAC throughregulating the TGF-120573 pathway and is critical to advancedtumors [12]

Further heredofamilial pancreatic adenocarcinoma isassociated with certain other significant genes with morecomplexmechanisms As indicated by the available literaturemost such genes participate in the DNA repair process suchas MSH12 PMS12 and BRCA12 which may participatein nonspecific tumor induction [13ndash15] However in severalinherited familial conditions specific known mutations havenot been identified which may hint at the complexity ofcarcinogenicmechanisms and the potential oncogenes aswellas tumor suppressor genes [16]

To predict more potential tumor-related genes we pro-posed a new method that considers protein interactionsfrom STRING (Search Tool for the Retrieval of InteractingGenesProteins) [17] and mines potential PC-related genesSTRING is a database with massive amounts of informationon physical and functional associations between differentproteinsWith an established score system STRINGenables auser to search and browse the protein interactions data as wellas simultaneously quantify the statistic cooccurrence in thebackground [17] Here depending on the method and usinga comprehensive analysis of the protein interaction networkwe fully utilized the database containing reported PC-relatedgenes and predicted potential genes involved in PC

2 Materials and Methods

21 Materials PC-related genes were retrieved from KEGGPATHWAY which is one of the main databases in KEGG(Kyoto Encyclopedia of Genes and Genomes) [18] By exam-ining the pathway hsa05212 pancreatic cancer (httpwwwgenomejpdbget-binwww bgethsa05212 accessed in De-cember 2014) we obtained 65 PC-related genes whichcomprise the gene set S Detailed information on these genesis listed in Supplementary Material I available online athttpdxdoiorg1011552015623121

22 Method for Mining New Candidate Genes in a Protein-Protein Interaction Network Protein-protein interaction(PPI) information is useful for investigating protein-relatedand gene-related problems [19ndash25] Most methods forpredicting protein attributes are based on the notion thattwo proteins that interact always share similar attributes[19ndash23] Because PC-related genes must share commonfeatures related to PC it is reasonable to use PPI informationto identify whether a gene is related to PC Here we adoptedthe PPI information reported in STRING (version 91httpwwwstring-dborg) [17] which is a large onlinedatabase that reports known and predicted protein interac-tions The PPIs reported in STRING are derived from thefollowing four sources (1) genomic context (2) high-throughput experiments (3) (conserved) coexpression and(4) previous knowledge which imply that the physical andfunctional associations of the proteins were measured Toextract PPI information for humans we downloaded the file

ldquoproteinlinksv91txtgzrdquo and selected protein interactionsbeginning with ldquo9606rdquo which produced 2425314 humanPPIs Each PPI includes two proteins and one score whichmeasures the strength of the interaction in the range between150 and 999 For the formulation let us denote the score ofan interaction between proteins 119901

1and 119901

2as 119876119894(1199011 1199012)

The aforementioned PPIs were used to construct a largenetwork by taking proteins as nodes Two nodes are adjacentif and only if the corresponding proteins comprise an inter-action that is a member of the 2425314 PPIs Furthermorethe interaction score should also be added to the constructednetwork To generate compatibility between our network anda shortest path approach each edge e was assigned a weight119908(119890) as follows 119908(119890) = 1000 minus 119876

119894(1199011 1199012) where 119901

1and 119901

2

are corresponding proteins of the endpoints of 119890Two proteins that can interact with each other always

share similar attributes [19ndash23] Further considering theinteraction score this notion can be generalized as followstwo proteins in an interaction with a high score are morelikely to share similar attributes than those with a low scoreMoreover if we consider a series of proteins 119901

1 1199012 119901

119899

such that two consecutive proteins can comprise an inter-action with a high score then these proteins may all sharesome common attributes By mapping these proteins ontothe constructed network they may comprise a shortest pathconnecting 119901

1and 119901

119899 which is a path connecting 119901

1and 119901

119899

such that the summation of the weights of edges on the pathis minimumThus we searched all shortest paths connectingany two PC-related genes For two consecutive nodes in eachof these paths their corresponding proteins can comprise aPPI with a high score because they lie on a shortest pathAs mentioned above they can share similar functions Fora specific shortest path 119901

1 1199012 119901

119899 where 119901

1and 119901

119899are

encoded byPC-related genes1199012shares similar functionswith

1199011 that is 119901

2may be encoded by an invalidated or known

PC-related gene 1199013shares similar functions with 119901

2 thus 119901

3

may also share similar functions with 1199011andmay be encoded

by an invalidated or known PC-related gene This can beinduced to 119901

4 1199015 119901

119899minus1 Thus we extracted genes in these

shortest paths and excluded those that were members of 65PC-related genes These genes were referred to as shortestpath genes and deemed to have special relationships with PCThe similar scheme has been applied to extract novel genesor chemicals related to other diseases or some biologicalprocesses [26ndash29] In addition to distinguish those geneseach shortest path genewas assigned a value which is referredto as betweenness and defined as the number of shortest pathsthat contain the gene The betweenness indicates the directand indirect associations between shortest path genes andPC-related genes [30]

23 Further Selection By executing themethodmentioned inSection 22 certain shortest path genes can be extracted fromthe constructed network However certain such genes maybe false positives To exclude this type of gene the followingmethod was adopted

Certain nodes in the constructed network are generalhubs their corresponding genes may always receive a high

BioMed Research International 3

betweenness value even if we randomly selected certain genesas PC-related genes but the genes exhibit few relationshipswith PC To exclude this type of gene among shortest pathgenes we randomly produced 1000 gene sets with the samesize as S and compared the betweenness of the shortest pathgenes in these sets to SThe detailed procedures are describedas follows

(I) Randomly produce 1000 gene sets such as 1198781

1198782 119878

1000 each of which with the same size as S

the set consisting of PC-related genes

(II) Themethod described in Section 22 was executed for1000 rounds For the 119894th round the PC-related genesin S were replaced with the genes in 119878

119894 and the

shortest paths connecting any pair of genes in 119878119894were

searched thereby counting the betweenness of eachshortest path gene based on these shortest paths

(III) For each shortest path gene there were 1000betweenness on randomly produced gene sets andone betweenness on S After comparing them wegenerated a measurement referred to as permutationFDR which defines the proportion of randomlyproduced gene sets on which the betweenness waslarger than that on S

(IV) To exclude shortest path genes with general hubsin the network and few relationships with PC weexcluded shortest path genes with permutation FDRvalues equal to or larger than 005 The remainingshortest path genes are referred to as candidate genes

3 Results and Discussion

31 Shortest Path Genes Based on the method described inSection 22 we searched all shortest paths connecting anypair of PC-related genes and produced 2080 shortest paths(each pair of PC-related genes can be connected by a shortestpath) which are provided in Supplementary Material II Agraph with these 2080 shortest paths is shown in Figure 1The detailed information of edges in this graph is providedin Supplementary Material III These 2080 paths involved134 Ensembl gene IDs We excluded 65 Ensembl IDs for PC-related genes resulting in 69 shortest path genes which arelisted in Supplementary Material IV In addition we countedthe betweenness of each shortest path gene which is alsoprovided in Supplementary Material IV

32 Candidate Genes As mentioned in Section 31 severalshortest path genes were retrieved According to the principleunderlying the method in Section 22 these genes may havespecial relationships with PC However certain such genesmay be false positives and have few relationships with PCThus we performed a permutation test to control for thistype of genes After calculating the permutation FDR foreach shortest path gene which are listed in SupplementaryMaterial IV we discarded the genes with a permutation FDRgreater than or equal to 005 thereby generating thirteencandidate genes which are listed in Table 1

33 Analysis of Significant Candidate Genes Using ourmethod we predicted thirteen genes that may participatein PC Based on the principle underlying our method thecandidate genes are specifically connected with PC-relatedgenes Moreover such genes have all been reported as genesthat are relevant to PC and may exhibit diverse functions intumor initiation and invasion especially in PC

Among the candidate genes three are inhibitory genesthat may participate in PC through their respective mech-anisms The first candidate gene is NFKBIA (see row 1of Table 1 with betweenness 73 and permutation FDR0002) which is expressed in pancreatic tissues [31] NFKBIAencodes amember of theNF-kappa-B inhibitor family whichhas been confirmed to further participate in interactions withREL dimers to inhibit the NF-kappa-B pathway in processesof inflammation immune response and tumorigenesis [3233] Targeted by a specific microRNA miR196a NFKBIAhas been proved to be associated with PC especially in themetastasis process [31] As we havementioned above inmosttumors including pancreatic cancer NF-kappa-B pathwayhas been widely reported to be overactivated and has aclose connection with the patientsrsquo prognosis indicating theunderlying relationship between NFKBIA and the pancreaticcancer [34ndash37] Mutations have been widely reported tocontribute to specific functional alteration of crucial proteinsincluding NFKBIA in diseases especially in various cancersubtypes like pancreatic cancer [38ndash41] What is more asone of the crucial inhibitory components of NF-kappa-Bpathway NFKBIA has been reported to be downregulatedin cancer [42 43] The overexpression of NFKBIA has alsobeen reported to be associated with a better prognosis ofvarious treatment methods in different human tumor sub-types especially for the prognosis of patients that have takenalpha 1-adrenoceptor antagonists and radiotherapy [44 45]Further considering the inflammation associated functionof such gene such mutations and expression alteration maycontribute to the process of tumorigenesis through twoindividual regulation mechanisms proliferation associatedpathways that involve NF-kappa-B and specific immuneresponse associated pathways in tumor microenvironment[46] In addition to the NF-kappa-B pathway the JNKpathway is also a specific pathway in tumor initiation andprogression including PC [47 48] A specific regulator ofRho protein exchange reactions which is crucial for JNKpathway ARHGDIA (see row 2 of Table 1 with between-ness 94 and permutation FDR 0012) was also identifiedusing our method and functions in several types of tumors[49 50] In addition this apoptosis inhibitory protein hasbeen confirmed to control Rho protein homeostasis andparticipate in the initiation and progression of PC throughthe apoptosis associated GDPGTP exchange reaction viaRhoA-Rho pathway [51] Apoptosis inhibition widely existsin tumor tissues and is an effective way to induce malignantcells implying the potential relationship between ARHGDIAand pancreatic cancer [52 53] As we all know pathwaysthat contribute to the proliferation and invasion of cancercells (JNK-STAT RhoA-Rho signaling pathway etc) haveall been confirmed to involve phosphorylation and dephos-phorylation process which may be regulated by the survival


3

6

1

1

3

1

2

1

1

1

11

1

1

1

11

1

1

1

1

3

1

10

2

6

1

1

1

11

5

2

1

2

1

1

1

1

1

3

1

1

1

1

1

1

1

1

1

1

1

21

1

11

1

1

2

1

2

1

2

1

1

1

1

13

6

8

1

1

1

2

1

1

1

1

11

1

1

2

1

1

1

6

5

1

1

1

1

11

33

1

1

4

1

2

1

1

1

1

1

11

2

5

1

1

1

3

2

1

1

1 1 14

2

2

1

2

1

1

1

2

1

1

1

1

1

4

2

2

3

3

1

28

6

1

1

1

1

1

1

2

1 13

1

1

11

1

1

1

1

11

1

11 1

1

1

3

1 11

1

3

1

11

21

11 1

2

1

1

1

1 1

1

2

1

1

1

1

11

11

111

1

1

1 1 1

1

1 1 1

1

1

1

1

1

1

3

1

1

1

2

2

1

11

1

1

1 1

1

111

1

35

1

1

1

1

11

1

1

1

1

2

2 11

1 1

1

1

1 11 11

1

11

2

1

1

1

11

ENSP00000282561

ENSP00000361120

ENSP00000295400ENSP00000366563

ENSP00000384675

ENSP00000278568

ENSP00000287647

ENSP00000293288ENSP00000366244

ENSP00000340944

ENSP00000019317

ENSP00000355896

ENSP00000361125

ENSP00000005257

ENSP00000375892

ENSP00000330237

ENSP00000329623

ENSP00000269300ENSP00000352121ENSP00000321410

ENSP00000268182ENSP00000278616

ENSP00000302564ENSP00000369497

ENSP00000362649

ENSP00000355249ENSP00000269571ENSP00000345571

ENSP00000266970ENSP00000361423

ENSP00000265171

ENSP00000267101

ENSP00000267868

ENSP00000417281

ENSP00000284384

ENSP00000262904ENSP00000262741

ENSP00000288986

ENSP00000320940

ENSP00000262435ENSP00000329357

ENSP00000332973

ENSP00000298316

ENSP00000347858

ENSP00000371067

ENSP00000358022

ENSP00000003084

ENSP00000250894

ENSP00000269321ENSP00000288602

ENSP00000355153

ENSP00000274335

ENSP00000384515

ENSP00000338018

ENSP00000348461

ENSP00000338934

ENSP00000343204ENSP00000263025

ENSP00000304895

ENSP00000303830ENSP00000348986

ENSP00000289153

ENSP00000297494

ENSP00000309103

ENSP00000304283

ENSP00000262160

ENSP00000254066ENSP00000354394

ENSP00000309845

ENSP00000268058

ENSP00000341551

ENSP00000244007

ENSP00000046794

ENSP00000302269

ENSP00000269305

ENSP00000215832

ENSP00000302486ENSP00000226574

ENSP00000251849ENSP00000264033

ENSP00000353483ENSP00000228872ENSP00000352157

ENSP00000350941

ENSP00000162330

ENSP00000228307

ENSP00000360683ENSP00000206249

ENSP00000339151ENSP00000358622

ENSP00000262367

ENSP00000270202ENSP00000257904ENSP00000221930

ENSP00000359424ENSP00000384273

ENSP00000263826ENSP00000364133

ENSP00000222254ENSP00000314458

ENSP00000263967ENSP00000351905

ENSP00000360266ENSP00000309503

ENSP00000354558ENSP00000300161ENSP00000219476

ENSP00000335153ENSP00000216797ENSP00000350283

ENSP00000263253

ENSP00000272519

ENSP00000262613

ENSP00000227507

ENSP00000342793ENSP00000344818

ENSP00000349467

ENSP00000265734

ENSP00000222005ENSP00000261799

ENSP00000267163ENSP00000238682

ENSP00000264657

ENSP00000223023

ENSP00000249071

ENSP00000341189

ENSP00000299421

ENSP00000244741

ENSP00000339007

ENSP00000250617

ENSP00000344456

ENSP00000359206

ENSP00000275493ENSP00000300574

Figure 1 A graph consisting of 2080 shortest pathsThe nodes on the inner circle (red nodes) represent 65 PC-related genes while the nodeson the outer circle (blue nodes) represent 69 shortest path genes The numbers on the edges represent the weights of the edges

associated GDPGTP exchange reaction further validatingthe potential relationship betweenARHGDIA and pancreaticcancer [54ndash58]What ismore regarded as a therapeutic targetof pancreatic cancer ARHGDIA (also known asRhoGDI) hasbeen actually proved to be associated with the proliferationand invasion process of the pancreatic tumorigenesis whichmay further affect the prognosis of various patients [51]The expression quantity of such gene ARHGDIA as thespecific negative regulator of Rho protein exchange reactionshas been widely confirmed to be downregulated in tumorswhich further induce the activation of survival associatedRho GTPases and promote the progression of various tumors[55ndash57] As for specific mutations that may affect the bio-logical function of such gene coincidentally specific tripleY156FS101AS174A-RhoGDI mutation has been confirmed

to be related to the progression of pancreatic cancer [58]Furthermore we also obtained the gene XIAP as one ofthe candidate genes in our list (see row 3 of Table 1 withbetweenness 64 and permutation FDR 0006) that encodes afunctional apoptosis suppressor protein As the most potentapoptosis suppressor XIAP interacts with several caspasesdirectly and prevents the apoptosis process [59ndash61] In PCXIAP has been proved to be critical for the progressionand prognosis of this disease via several classical pathways(Erk PTENPI3KAKT etc) [62ndash64] What is more theoverexpression of XIAP has been confirmed to be associatedwith poor prognosis of pancreatic cancer Therefore suchprotein has already been applied as a useful laboratory testparameter and a functional therapeutic target of pancreaticcancer [64 65] As an immune associated functional gene


Table 1 Detailed information on the thirteen candidate genes

Row number Ensembl ID Gene symbol Full name Betweenness Permutation FDR

1 ENSP00000216797 NFKBIA Nuclear Factor of Kappa Light PolypeptideGene Enhancer in B-Cells Inhibitor Alpha 73 0002

2 ENSP00000269321 ARHGDIA Rho GDP Dissociation Inhibitor (GDI)Alpha 94 0012

3 ENSP00000347858 XIAP X-Linked Inhibitor of Apoptosis E3Ubiquitin Protein Ligase 64 0006

4 ENSP00000250894 MAPK8IP3 Mitogen-Activated Protein Kinase 8Interacting Protein 3 64 lt0001

5 ENSP00000262613 SLC9A3R1Solute Carrier Family 9 Subfamily A(NHE3 Cation Proton Antiporter 3)Member 3 Regulator 1

128 0031

6 ENSP00000384515 PARVB Parvin Beta 64 lt00017 ENSP00000254066 RARA Retinoic Acid Receptor Alpha 24 00058 ENSP00000299421 ILK1 Integrin-linked kinase 1 64 0039 ENSP00000338934 EZR Ezrin 128 0026

10 ENSP00000309845 HRAS Harvey Rat Sarcoma Viral OncogeneHomolog 113 0033

11 ENSP00000335153 HSP90AA1 Heat Shock Protein 90 kDa Alpha(Cytosolic) Class A Member 1 230 0007

12 ENSP00000348986 INS-IGF2 INS-IGF2 Readthrough 64 003113 ENSP00000284384 PRKCA Protein Kinase C Alpha 175 lt0001

the polymorphism of XIAP may contribute to the complexand personalized immune response to pancreatic cancersSome of the functional variants have already been proved tobe directly associated with the prognosis of various subtypesof pancreatic cancer indicating the potential significant roleof XIAP in pancreatic cancer [62 66]

In addition to inhibitory genes a group of scaffold geneswas also among the thirteen candidate genes MAPK8IP3(see row 4 of Table 1 with betweenness 64 and permutationFDR lt0001) which is also known as JIP3 is a scaffoldprotein involved in the JNK pathway [67] Associated withthe MAP kinase cascade it may be related to PC viapotential downstreammechanisms [67ndash69]With high tissuespecificity such gene has been reported to actually inducethe excessive proliferation of pancreatic cells which mayfurther initiate the biological process of pancreatic cancer[70]MAPK8IP3 is also one of the crucial components of JNKpathway which has been overactivated in tumors includingpancreatic tumor [71 72] During the initiation and invasionstages of pancreatic cancer MAPK8IP3 has been reportedto be overexpressed which may further contribute to theexcessive proliferation of pancreatic cells in clinical cases[70] Mutations have also been identified in MAPK8IP3which has been reported to partially interfere with its ownfunctions under pathological conditions [73 74] Some ofthe specific mutations of MAPK8IP3 have been reportedto influence the adhesion and invasion process of variouscancer cells including the pancreatic cancer cells validatingthe crucial role of MAPK8IP3 in pancreatic cancer [75 76]As another functional scaffold protein that participates inthe interaction between the cell membrane and cytoskeleton

our candidate gene SLC9A3R1 (see row 5 of Table 1 withbetweenness 128 and permutation FDR 0031) regulates thesubcellular location and function of SLC9A3 and plays aspecific role in theWnt pathway [77ndash79] Known asNHERF1SLC9A3R1 acts as a scaffold protein in many types of tumorsdepending on its specific role in Wnt pathway especiallyin PC [78 80 81] During the course of pancreatic cancerraising SLC9A3R1 has been confirmed to be overexpressedand involve the poor prognosis of such disease [81 82] Justlike other candidates encoding scaffold proteins SLC9A3R1has also been proved to contain specific nonsynonymousmutations in various cancer subtypes especially in pancreaticcancer [83 84] Coding an adaptor protein the candidategene PARVB (see row 6 of Table 1 with betweenness 64and permutation FDR lt0001) is critical to certain specificbiological processes in cancer such as ERK signaling pathwayand focal adhesion which both have been identified inpancreatic cancer [85 86] Therefore this gene is criticalfor tumorigenesis and may have a specific role in PC [87]As a crucial adaptor protein the overexpression of PARVBhas been proved to be associated with high cell proliferationrate which indicates worsening clinical prognosis of tumorsincluding pancreatic cancer [86 88] The polymerase of suchgene PARVB has also been identified and confirmed to beassociated with a group of severe diseases including cancers[87] Such abnormal expression level and high polymerase intumor cells suggest that PARVBmay play a crucial role in thespecific pathogenic process of pancreatic cancer

Six remaining genes may also be associated with PCin their respective way RARA (see row 7 of Table 1with betweenness 24 and permutation FDR 0005) encodes


a specific receptor for retinoic acid which may further reg-ulate germ cell development during spermatogenesis and theexpression of specific genes through recruiting the chromatincomplex (KMT2E MLL5 etc) [89ndash91] As we all know thestem-cell-like features of tumor cells may be associated withthe potential invasion and proliferation ability of such tumorcells [92ndash94] In pancreatic cancer it has been confirmed thatthe more ldquostem-likerdquo the cancer cell is the more malignantthe cancer is [95] Similar to its specific function in germ cellsRARA is related to the regulation of stem- cell-like features inPC which implies that it may further affect the prognosis ofsuch a severe malignancy [96] The overexpression of RARAhas been reported to be related to specific biological behav-iors of tumors such as EMT (epithelial-to-mesenchymal)and invasion which indicates poor prognosis of pancreaticcancers in clinical cases [97] Although few mutations havebeen identified in RARA crucial fusion gene variants havebeenwidely identified in cancerswhichmay be analyzed laterAs we have mentioned above SLC9A3R1 has been reportedto regulate cellular morphology as a scaffold protein [78]Similarly the candidate gene ILK1 (see row 8 of Table 1 withbetweenness 64 and permutation FDR 0005) also mediatescell architecture signaling transduction and cell adhesionvia integrin-mediated signaling transduction especially intumor tissues [98 99] Although studies have not explainedclearly how ILK1 actually contributes to the process of pancre-atic cancer it has been shown to participate in several types oftumors and may be associated with several unique molecules(GSK AKT PTEN etc) that all have been confirmed to becritical to PC [98ndash103] Recently such gene has been reportedto be associated with the expression of TGF-120572 which maypartially explain the underlying carcinogenic mechanism ofsuch gene [104]What is more such gene has been confirmedto be overexpressed during the whole biological process oftumorigenesis including the proliferation migration andinvasion of pancreatic cancers [104ndash106] Further specificmutations have also been identified in such gene Integrin-linked kinase 1 (ILK1) and HSP90 (HSP90AA1 which isalso in our list and will be analyzed later) have both beenconfirmed to contain quite a lot of tumor-associated variantsSuch mutations of ILK1 have been reported to be functionalpancreatic cancer biomarkers [107ndash109] All in all ILK1has been proved to regulate the tumorigenesis process ofPC through a specific TGF-120572-associated mechanism As afunctional SLC9A3R1-associated protein a specific proteinezrin is encoded by one of our candidate genes EZR (seerow 9 of Table 1 with betweenness 128 and permutation FDR0026) [110] Crucial for cytoplasmic peripheral membraneEZR is associated with the actin cytoskeleton and regulatesthe surface expression of actin [111 112] It has been widelyreported that ezrin is critical for the proliferation metastasisand invasion process of several tumor subtypes especiallyof pancreatic cancer [113 114] The overexpression of EZRhas been widely reported to be associated with the initia-tion proliferation and metastasis processes of tumor whichstrongly affect the patientsrsquo prognosis especially in pancreaticcancer [115ndash117] EZR as the crucial scaffold protein whichmay interact with SLC9A3R1 that we have mentioned abovehas also been reported to contain specific mutations that may

contribute to the tumorigenesis process of various cancersubtypes including pancreatic cancer [118] All in all EZRmay be a potential pancreatic-cancer-associated gene andmay contribute to similar pathways with SL9A3R1 as we havementioned above [110]

Further we also predicted one of the most famousoncogenes HRAS (see row 10 of Table 1 with betweenness113 and permutation FDR 0033) which binds GDPGTPand exhibits intrinsic GTPase activity [119] Associated withseveral tumors and similar to its homologues (r-RAS k-Ras etc) HRAS is a powerful oncogene that can initiatetumors through inducing excessive growth factor activationin cells and promoting malignant proliferation of tumorcells especially in tumors in situ [120ndash122] In PC suchgene has also been regarded as a main oncogene and iswidely reported to be overexpressed during the entire clinicalcourse of pancreatic cancer validating our prediction of thepotential pancreatic cancer-associated genes [123] Severalgenetic alterations in HRAS (mutations CNVs etc) havebeen identified and reported to be associated with such asevere type of cancer [124ndash126] Recently some of the crucialmutations have already been confirmed to directly con-tribute to the tumorigenesis of pancreatic cancer indicatingthe underlying relationship between mutations of candidategenes and the process of tumorigenesis [127] Heat shockproteins compose a unique group of proteins produced bycells to resist harmful and stressful conditions includingthe tumor microenvironment which contains less oxygenand a lower pH [128ndash130] This cluster of proteins has beenconfirmed to be functional in tumor microenvironment andpromote the progression of various tumor subtypes [131 132]One of the candidate genes HSP90AA1 (see row 11 of Table 1with betweenness 230 and permutation FDR 0007) alsoencodes a specific heat shock protein heat shock protein90 kDa alpha which has been reported to be expressed inthe cytoplasm Several types of heat shock proteins (suchas HSP20 HSP70 and HSP90) including our predictedprotein are not only related to tumor genesis but specificallycontribute to PC [133ndash135] Heat shock proteins have alsobeen confirmed to be overexpressed in pancreatic cancerincluding HSP90 which is encoded by our predicted gene[136ndash140] The overexpression of heat shock proteins hasbeen confirmed to actually contribute to the tumorigenesisprocess of pancreatic cancer which validates our predicationof HSP90AA1 as a functional candidate gene [133] As ahormone-associated gene our candidate gene INS-IGF2 (seerow 12 of Table 1 with betweenness 64 and permutationFDR 0031) is a readthrough gene of INS and IGF2 Thisgene may act as a posttranscriptional regulatory factor forthe two genes INS and IGF2 which has been shown to beassociated with several pancreatic-associated diseases suchas diabetes [141ndash143] As a specific fusion gene INS-IGF2participates in several cancer types especially in prostatecancer [144ndash146] The fusion of a tumor-associated geneand a pancreas-related gene INS-IGF2 probably participatesin PC progression and invasion specifically Recently suchfusion gene INS-IGF2 is reported to be specifically expressedin pancreatic islets especially in patients with autoimmunediseases under pathological conditions [147 148] Since


immune response is quite significant for tumor surveillanceINS-IGF2 is definitely associated with the unique tumor-associated immune response and may further contribute tothe prognosis of pancreatic cancer [148] Although genes likeRARA (mostly identified as FIP1L1PML-RARA fusion gene)(we have mentioned above) and INS-IGF2 have not beenreported to contain specific mutations that may be associatedwith pancreatic cancer such fusion genes have both been con-firmed to have abnormal gene functions which may directlycontribute to the tumorigenesis of various cancer subtypesincluding pancreatic cancer [149 150] The last candidategene PRKCA (see row 13 of Table 1 with betweenness 175and permutation FDR lt0001) encodes a specific kinasethat has been reported in several types of tumors [151 152]As a widely reported oncogene PRKCA contributes to thephosphorylation process of crucial tumor-associated proteins[153] The overexpression of PRKCA has been proved toenhance the transformed proliferation and invasion processof pancreatic cancer which further definitely influences theprognosis of such disease [154 155] While at the sametime genes like PRKCA which encodes functional kinasemay be strongly affected by genetic alterations especially bysome specific nonsynonymousmutations [75 156]Thereforethe polymerase of gene PRKCA may definitely affect thefunction of protein kinase C which has also been identifiedin pancreatic cancer [156 157] However recent studiesshow that at least in vitro PRKCA has a specific functionin tumor suppression [156] The complex role of PKC incancer may reflect the complicated interactive relationshipbetween PRKCA and PC [156 158] All in all our predictedgene PRKCA has been confirmed to be definitely associatedwith pancreatic cancer while at the same time the variousfunctions of PRKCA imply the complex role of such gene inpancreatic cancer

Consequently all candidate genes obtained using ourmethod play a critical role in tumor initiation and progres-sion Because the detailed regulationmechanisms of only fewgenes remain unclear those candidate genes may be genesthat are critical to PC PC features genetic heterogeneityand many familial hereditary related genes remain ambigu-ous [16] Therefore in addition to the confirmed genessuch as KRAS and p16INK41 [8 9] more genes must bescreened and validated to clearly demonstrate the mecha-nisms of pancreatic adenocarcinoma In conclusion basedon reported PC-related genes the proposed method waseffective at predicting candidate tumor-associated genes inPC

4 Conclusions

In this study we proposed a computational method to minenew candidate genes related to pancreatic cancer whichutilized protein-protein interaction information The analy-ses of the obtained candidate genes indicate that they maybe novel PC-related genes Hopefully this contribution willpromote studies on pancreatic cancer and provide new hopefor designing effective treatments

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

This study was supported by the National Basic ResearchProgram of China (2011CB510101 and 2011CB510102) theNational Natural Science Foundation of China (31371335)and the Innovation Program of Shanghai Municipal Educa-tion Commission (12ZZ087)

References

[1] M Hidalgo S Cascinu J Kleeff et al ldquoAddressing the chal-lenges of pancreatic cancer future directions for improvingoutcomesrdquo Pancreatology vol 15 no 1 pp 8ndash18 2015

[2] C Verbeke M Lohr J Severin Karlsson and M Del ChiaroldquoPathology reporting of pancreatic cancer following neoadju-vant therapy challenges and uncertaintiesrdquo Cancer TreatmentReviews vol 41 no 1 pp 17ndash26 2015

[3] J Freilich T Strom G Springett et al ldquoAge and resectedpancreatic cancer outcomesrdquo International Journal of RadiationOncology Biology Physics vol 90 no 1 supplement p S3562014

[4] S Mizuno Y Nakai H Isayama et al ldquoSmoking family historyof cancer and diabetes mellitus are associated with the age ofonset of pancreatic cancer in japanese patientsrdquo Pancreas vol43 no 7 pp 1014ndash1017 2014

[5] G Preziosi J A Oben and G Fusai ldquoObesity and pancreaticcancerrdquo Surgical Oncology vol 23 no 2 pp 61ndash71 2014

[6] S Deng S Zhu B Wang et al ldquoChronic pancreatitis andpancreatic cancer demonstrate active epithelial-mesenchymaltransition profile regulated by miR-217-SIRT1 pathwayrdquo CancerLetters vol 355 no 2 pp 184ndash191 2014

[7] K S Snima R S Nair S V Nair C R Kamath and V-KLakshmanan ldquoCombination of anti-diabetic drug metforminand boswellic acid nanoparticles a novel strategy for pancreaticcancer therapyrdquo Journal of Biomedical Nanotechnology vol 11no 1 pp 93ndash104 2015

[8] J Luttges A Reinecke-Luthge B Mollmann et al ldquoDuctchanges and K-ras mutations in the disease-free pancreasanalysis of type age relation and spatial distributionrdquo VirchowsArchiv vol 435 no 5 pp 461ndash468 1999

[9] D Georgiadou T N Sergentanis S Sakellariou et al ldquoCyclinD1 p16INK4A and p27Kip1 in pancreatic adenocarcinoma assess-ing prognostic implications through quantitative image analy-sisrdquo Acta Pathologica Microbiologica Et Immunologica Scandi-navica vol 122 no 12 pp 1230ndash1239 2014

[10] Z Wang S Ali S Banerjee et al ldquoActivated K-Ras andINK4aArf deficiency promote aggressiveness of pancreaticcancer by induction of EMT consistent with cancer stem cellphenotyperdquo Journal of Cellular Physiology vol 228 no 3 pp556ndash562 2013

[11] L Zhang W Liu Y Qin and R Wu ldquoExpression of TGF-1205731in Wilmsrsquo tumor was associated with invasiveness and diseaseprogressionrdquo Journal of Pediatric Urology vol 10 no 5 pp 962ndash968 2014


[12] Y El-Gohary S Tulachan P Guo et al ldquoSmad signalingpathways regulate pancreatic endocrine developmentrdquoDevelop-mental Biology vol 378 no 2 pp 83ndash93 2013

[13] J A Veenstra ldquoImmunocytochemical demonstration of ahomology in peptidergic neurosecretory cells in the suboe-sophageal ganglion of a beetle and a locust with antisera tobovine pancreatic polypeptide FMRFamide vasopressin and120572-MSHrdquo Neuroscience Letters vol 48 no 2 pp 185ndash190 1984

[14] R Das Gupta and R Kolodner ldquoIdentification of postmeioticsegregation (PMS) mutants as novel alleles of mismatch repair(MMR) genesrdquoThe FASEB Journal vol 11 no 9 p 1971 1997

[15] T Golan Z S Kanji R Epelbaum et al ldquoOverall survival andclinical characteristics of pancreatic cancer in BRCA mutationcarriersrdquo British Journal of Cancer vol 111 no 6 pp 1132ndash11382014

[16] A P Klein K A Brune G M Petersen et al ldquoProspectiverisk of pancreatic cancer in familial pancreatic cancer kindredsrdquoCancer Research vol 64 no 7 pp 2634ndash2638 2004

[17] D Szklarczyk A Franceschini S Wyder et al ldquoSTRING v10protein-protein interaction networks integrated over the treeof liferdquo Nucleic Acids Research vol 43 no 1 pp D447ndashD4522015

[18] M Kanehisa and S Goto ldquoKEGG kyoto encyclopedia of genesand genomesrdquo Nucleic Acids Research vol 28 no 1 pp 27ndash302000

[19] M Deng K Zhang S Mehta T Chen and F Sun ldquoPredictionof protein function using protein-protein interaction datardquoJournal of Computational Biology vol 10 no 6 pp 947ndash9602003

[20] L Hu T Huang X Shi W-C Lu Y-D Cai and K-C ChouldquoPredicting functions of proteins in mouse based on weightedprotein-protein interaction network and protein hybrid prop-ertiesrdquo PLoS ONE vol 6 no 1 Article ID e14556 2011

[21] P Gao Q-P Wang L Chen and T Huang ldquoPrediction ofhuman genes regulatory functions based on proteinproteininteraction networkrdquo Protein and Peptide Letters vol 19 no 9pp 910ndash916 2012

[22] K-L Ng J-S Ciou and C-H Huang ldquoPrediction of proteinfunctions based on function-function correlation relationsrdquoComputers in Biology and Medicine vol 40 no 3 pp 300ndash3052010

[23] R Sharan I Ulitsky and R Shamir ldquoNetwork-based predictionof protein functionrdquoMolecular Systems Biology vol 3 article 882007

[24] P Bogdanov and A K Singh ldquoMolecular function predictionusing neighborhood featuresrdquo IEEEACMTransactions onCom-putational Biology and Bioinformatics vol 7 no 2 pp 208ndash2172010

[25] M D M AbdulHameed G J Tawa K Kumar et al ldquoSystemslevel analysis and identification of pathways and networksassociated with liver fibrosisrdquo PLoS ONE vol 9 no 11 ArticleID e112193 2014

[26] L Chen C Chu J Lu X Kong T Huang and Y D CaildquoA computational method for the identification of new candi-date carcinogenic and non-carcinogenic chemicalsrdquo MolecularBioSystems vol 11 no 9 pp 2541ndash2550 2015

[27] T Gui X Dong R Li Y Li and Z Wang ldquoIdentification ofhepatocellular carcinoma-related genes with a machine learn-ing and network analysisrdquo Journal of Computational Biologyvol 22 no 1 pp 63ndash71 2015

[28] L Chen C Chu X Kong G Huang T Huang and Y CaildquoA hybrid computational method for the discovery of novelreproduction-related genesrdquo PLoS ONE vol 10 no 3 ArticleID e0117090 2015

[29] B-Q Li J Zhang T Huang L Zhang and Y-D Cai ldquoIdenti-fication of retinoblastoma related genes with shortest path in aprotein-protein interaction networkrdquo Biochimie vol 94 no 9pp 1910ndash1917 2012

[30] J B M Craven Markov Networks for Detecting OverlappingElements in Sequence Data MIT Press 2005

[31] F Huang J Tang X Zhuang et al ldquoMiR-196a promotes pan-creatic cancer progression by targeting nuclear factor kappa-B-inhibitor alphardquo PLoSONE vol 9 no 2 Article ID e87897 2014

[32] P-Y Chang K Draheim M A Kelliher and S MiyamotoldquoNFKB1 is a direct target of the TAL1 oncoprotein in human Tleukemia cellsrdquo Cancer Research vol 66 no 12 pp 6008ndash60132006

[33] N Fullard C L Wilson and F Oakley ldquoRoles of c-Relsignalling in inflammation and diseaserdquo International Journal ofBiochemistry and Cell Biology vol 44 no 6 pp 851ndash860 2012

[34] Y Ito E Kikuchi N Tanaka et al ldquoDown-regulation of NFkappa B activation is an effective therapeutic modality inacquired platinum-resistant bladder cancerrdquo BMC Cancer vol15 article 324 2015

[35] L Lessard A-M Mes-Masson L Lamarre L Wall J-B Lat-touf and F Saad ldquoNF-120581Bnuclear localization and its prognosticsignificance in prostate cancerrdquo BJU International vol 91 no 4pp 417ndash420 2003

[36] T Gilmore M-E Gapuzan D Kalaitzidis and D Star-czynowski ldquoRelNF-120581BI120581B signal transduction in the gener-ation and treatment of human cancerrdquo Cancer Letters vol 181no 1 pp 1ndash9 2002

[37] B Haefner ldquoNF-kappa B arresting a major culprit in cancerrdquoDrug Discovery Today vol 7 no 12 pp 653ndash663 2002

[38] H Kinugasa K Nouso K Miyahara et al ldquoDetection of K-ras gene mutation by liquid biopsy in patients with pancreaticcancerrdquo Cancer vol 121 no 13 pp 2271ndash2280 2015

[39] A Purohit M Varney S Rachagani M Ouellette S K Batraand R K Singh ldquoCXCR2 signaling axis- a helping hand toKRAS(G12D) mutation in pancreatic cancer initiationrdquo CancerResearch vol 74 article 4874 2014

[40] H Hayashi H Ueno Y Sakamoto et al ldquoGenemutation profileof pancreatic cancer in Japanese patients and its associationwithprognosisrdquo Annals of Oncology vol 25 2014

[41] K Saito Y Nakai H Isayama et al ldquoAssociations betweenK-rasmutation smoking and prognosis of pancreatic cancerrdquo Journalof Clinical Oncology vol 32 supplement 3 abstract 298 2014

[42] M Bredel H Kim T K Nanda et al ldquoDeletion of thetumor suppressor NFKBIA in triple-negative breast cancerrdquoInternational Journal of Radiation Oncology Biology Physicsvol 87 no 2 p S98 2013

[43] G R Harsh D M Scholtens A K Yadav et al ldquoNFKBIAin glioblastomas tumor suppressor and potent predictor ofoutcomerdquo Journal of Neurosurgery vol 115 p A404 2011

[44] T Mukogawa F Koyama M Tachibana et al ldquoAdenovirus-mediated gene transduction of truncated I120581B120572 enhancesradiosensitivity in human colon cancer cellsrdquo Cancer Sciencevol 94 no 8 pp 745ndash750 2003

[45] J V Partin I E Anglin and N Kyprianou ldquoQuinazoline-based 1205721-adrenoceptor antagonists induce prostate cancer cellapoptosis via TGF-120573 signalling and I120581B120572 inductionrdquo BritishJournal of Cancer vol 88 no 10 pp 1615ndash1621 2003


[46] K M Reid-Lombardo B L Fridley W R Bamlet J MCunningham M G Sarr and G M Petersen ldquoSurvival isassociated with genetic variation in inflammatory pathwaygenes among patients with resected and unresected pancreaticcancerrdquo Annals of Surgery vol 257 no 6 pp 1096ndash1102 2013

[47] S Suzuki M Okada K Shibuya et al ldquoJNK suppressionof chemotherapeutic agents-induced ROS confers chemoresis-tance on pancreatic cancer stem cellsrdquo Oncotarget vol 6 no 1pp 458ndash470 2015

[48] X P YuanM Dong X Li and J P Zhou ldquoGRP78 promotes theinvasion of pancreatic cancer cells by FAK and JNKrdquoMolecularand Cellular Biochemistry vol 398 no 1-2 pp 55ndash62 2014

[49] L Liang Q Li L YHuang et al ldquoLoss of ARHGDIA expressionis associated with poor prognosis in HCC and promotesinvasion and metastasis of HCC cellsrdquo International Journal ofOncology vol 45 no 2 pp 659ndash666 2014

[50] M Dal Bo F Pozzo R Bomben et al ldquoARHGDIA a mutantTP53-associated Rho GDP dissociation inhibitor is over-expressed in gene expression profiles of TP53 disrupted chroniclymphocytic leukaemia cellsrdquo British Journal of Haematologyvol 161 no 4 pp 596ndash599 2013

[51] M A Harding and D Theodorescu ldquoRhoGDI signaling pro-vides targets for cancer therapyrdquo European Journal of Cancervol 46 no 7 pp 1252ndash1259 2010

[52] D Dixit V Sharma S Ghosh V S Mehta and E SenldquoInhibition of Casein kinase-2 induces p53-dependent cellcycle arrest and sensitizes glioblastoma cells to tumor necrosisfactor (TNF120572)-induced apoptosis through SIRT1 inhibitionrdquoCell Death and Disease vol 3 no 2 article e271 2012

[53] Q Liu X-Y Zhao R-Z Bai et al ldquoInduction of tumorinhibition and apoptosis by a candidate tumor suppressor geneDRR1 on 3p211rdquoOncology Reports vol 22 no 5 pp 1069ndash10752009

[54] M J Marinissen M Chiariello T Tanos O Bernard S Naru-miya and J S Gutkind ldquoThe small GTP-binding protein RhoAregulates c-jun by a ROCK-JNK signaling axisrdquoMolecular Cellvol 14 no 1 pp 29ndash41 2004

[55] W P Bozza Y Zhang K Hallett L A Rivera Rosado and BZhang ldquoRhoGDI deficiency induces constitutive activation ofRho GTPases and COX-2 pathways in association with breastcancer progressionrdquo Oncotarget 2015

[56] S Hooshmand A Ghaderi K Yusoff KThilakavathy R Rosliand Z Mojtahedi ldquoDifferentially expressed proteins in ER+MCF7 and ERminus MDA-MB-231 human breast cancer cells byRhoGDI-120572 silencing and overexpressionrdquo Asian Pacific Journalof Cancer Prevention vol 15 no 7 pp 3311ndash3317 2014

[57] B Zhang ldquoRho GDP dissociation inhibitors as potential targetsfor anticancer treatmentrdquoDrug Resistance Updates vol 9 no 3pp 134ndash141 2006

[58] Z XWang and D CThurmond ldquoDifferential phosphorylationof RhoGDI mediates the distinct cycling of Cdc42 and Rac1 toregulate second-phase insulin secretionrdquo Journal of BiologicalChemistry vol 285 no 9 pp 6186ndash6197 2010

[59] G Jiao W Guo T Ren et al ldquoBMPR2 inhibition inducedapoptosis and autophagy via destabilization of XIAP in humanchondrosarcoma cellsrdquo Cell Death and Disease vol 5 no 12Article ID e1571 2014

[60] R Hu Y Yang Z Liu et al ldquoThe XIAP inhibitor Embelinenhances TRAIL-induced apoptosis in human leukemia cells byDR4 and DR5 upregulationrdquo Tumor Biology vol 36 no 2 pp769ndash777 2015

[61] Z Cao X Li J Li W Luo C Huang and J Chen ldquoX-linkedinhibitor of apoptosis protein (XIAP) lacking RING domainlocalizes to the nuclear and promotes cancer cell anchorage-independent growth by targeting the E2F1Cyclin E axisrdquoOncotarget vol 5 no 16 pp 7126ndash7137 2014

[62] X P Yi T Han Y X Li X Y Long andW Z Li ldquoSimultaneoussilencing of XIAP and survivin causes partial mesenchymal-epithelial transition of human pancreatic cancer cells via thePTENPI3KAkt pathwayrdquoMolecular Medicine Reports vol 12pp 601ndash608 2015

[63] L-P Cao J-L Song X-P Yi and Y-X Li ldquoDouble inhibitionof NF-120581B and XIAP via RNAi enhances the sensitivity ofpancreatic cancer cells to gemcitabinerdquo Oncology Reports vol29 no 4 pp 1659ndash1665 2013

[64] Z-H Wang H Chen H-C Guo et al ldquoEnhanced antitu-mor efficacy by the combination of emodin and gemcitabineagainst human pancreatic cancer cells via downregulation of theexpression of XIAP in vitro and in vivordquo International Journalof Oncology vol 39 no 5 pp 1123ndash1131 2011

[65] H Zai X Yi Y Li C Jiang and X Lu ldquoSimultaneous inhi-bition of XIAP and survivin expression on EMT and invasionof human pancreatic cancer cellsrdquo Journal of Central SouthUniversity vol 37 no 9 pp 883ndash888 2012

[66] R Yu SM Albarenque R H CoolW J Quax AMohr and RM Zwacka ldquoDR4 specific TRAIL variants are more efficaciousthanwild-type TRAIL in pancreatic cancerrdquoCancer Biology andTherapy vol 15 no 12 pp 1658ndash1666 2014

[67] T Sun N Yu L-K Zhai et al ldquoC-Jun NH2-terminal kinase(JNK)-interacting protein-3 (JIP3) regulates neuronal axonelongation in a kinesin- and JNK-dependent mannerrdquo Journalof Biological Chemistry vol 288 no 20 pp 14531ndash14543 2013

[68] H Matsuura H Nishitoh K Takeda et al ldquoPhosphorylation-dependent scaffolding role of JSAP1JIP3 in the ASK1-JNKsignaling pathway A newmode of regulation of theMAPkinasecascaderdquoThe Journal of Biological Chemistry vol 277 no 43 pp40703ndash40709 2002

[69] N Skrypek R Vasseur A Vincent B Duchene I VanSeuningen and N Jonckheere ldquoThe oncogenic receptor ErbB2modulates gemcitabine and irinotecanSN-38 chemoresistanceof human pancreatic cancer cells via hCNT1 transporter andmultidrug-resistance associated protein MRP-2rdquo Oncotargetvol 6 no 13 pp 10853ndash10867 2015

[70] J Stoslashrling N Allaman-Pillet A E Karlsen N Billestrup CBonny and T Mandrup-Poulsen ldquoAntitumorigenic effect ofproteasome inhibitors on insulinoma cellsrdquo Endocrinology vol146 no 4 pp 1718ndash1726 2005

[71] C L Standen N J Kennedy R A Flavell and R J DavisldquoSignal transduction cross talk mediated by Jun N-terminalkinase-interacting protein and insulin receptor substrate scaf-fold protein complexesrdquoMolecular and Cellular Biology vol 29no 17 pp 4831ndash4840 2009

[72] P P Ongusaha H H Qi L Raj et al ldquoIdentification of ROCK1as an upstream activator of the JIP-3 to JNK signaling axis inresponse to UVB damagerdquo Science Signaling vol 1 no 47 pra14 2008

[73] S N Rao and V N Balaji ldquoMolecular docking studies on JNKinhibitors at the allosteric JNK-JIP interaction siterdquo CurrentScience vol 102 no 7 pp 1009ndash1016 2012

[74] W Engstrom andM Granerus ldquoExpression of JNK-interactingprotein JIP-1 and insulin-like growth factor II inWilms tumourcell lines and primary Wilms tumoursrdquo Anticancer Researchvol 29 no 7 pp 2467ndash2472 2009


[75] T Takino M Nakada H Miyamori et al ldquoJSAP1JIP3 coop-erates with focal adhesion kinase to regulate c-Jun N-terminalkinase and cell migrationrdquoThe Journal of Biological Chemistryvol 280 no 45 pp 37772ndash37781 2005

[76] T Matsuguchi A Masuda K Sugimoto Y Nagai and YYoshikai ldquoJNK-interacting protein 3 associates with Toll-likereceptor 4 and is involved in LPS-mediated JNK activationrdquoTheEMBO Journal vol 22 no 17 pp 4455ndash4464 2003

[77] N Broere M Chen A Cinar et al ldquoDefective jejunal andcolonic salt absorption and alteredNa+H+ exchanger 3 (NHE3)activity in NHE regulatory factor 1 (NHERF1) adaptor protein-deficient micerdquo Pflugers ArchivmdashEuropean Journal of Physiol-ogy vol 457 no 5 pp 1079ndash1091 2009

[78] A Bellizzi M R Greco R Rubino et al ldquoThe scaffoldingprotein NHERF1 sensitizes EGFR-dependent tumor growthmotility and invadopodia function to gefitinib treatment inbreast cancer cellsrdquo International Journal of Oncology vol 46no 3 pp 1214ndash1224 2015

[79] B Hopwood A Tsykin D M Findlay and N L FazzalarildquoMicroarray gene expression profiling of osteoarthritic bonesuggests altered bone remodelling WNT and transforminggrowth factor-120573bone morphogenic protein signallingrdquo Arthri-tis Research andTherapy vol 9 no 5 article R100 2007

[80] Y Jiang S Wang J Holcomb et al ldquoCrystallographic analysisof NHERF1-PLCbeta3 interaction provides structural basisfor CXCR2 signaling in pancreatic cancerrdquo Biochemical andBiophysical Research Communications vol 446 no 2 pp 638ndash643 2014

[81] SWang YWu YHou et al ldquoCXCR2macromolecular complexin pancreatic cancer a potential therapeutic target in tumorgrowthrdquo Translational Oncology vol 6 no 2 pp 216ndash225 2013

[82] A Mangia C Saponaro A Malfettone et al ldquoInvolvement ofnuclear NHERF1 in colorectal cancer progressionrdquo OncologyReports vol 28 no 3 pp 889ndash894 2012

[83] Y Hayashi J R Molina S R Hamilton and M-M GeorgesculdquoNHERF1EBP50 is a newmarker in colorectal cancerrdquoNeopla-sia vol 12 no 12 pp 1013ndash1022 2010

[84] Y Y Jiang S O Wang J Holcomb et al ldquoCrystallographicanalysis of NHERF1-PLC1205733 interaction provides structuralbasis for CXCR2 signaling in pancreatic cancerrdquo Biochemicaland Biophysical Research Communications vol 446 no 2 pp638ndash643 2014

[85] T Fukuda L Guo X Shi and CWu ldquoCH-ILKBP regulates cellsurvival by facilitating the membrane translocation of proteinkinase BAktrdquo The Journal of Cell Biology vol 160 no 7 pp1001ndash1008 2003

[86] A Eslami K Miyaguchi K Mogushi et al ldquoPARVB overex-pression increases cellmigration capability anddefines high riskfor endophytic growth and metastasis in tongue squamous cellcarcinomardquo British Journal of Cancer vol 112 no 2 pp 338ndash344 2015

[87] J L Sepulveda andCWu ldquoTheparvinsrdquoCellular andMolecularLife Sciences vol 63 no 1 pp 25ndash35 2006

[88] G Rosenberger I Jantke A Gal and K Kutsche ldquoInteractionof 120572PIX (ARHGEF6) with 120573-parvin (PARVB) suggests aninvolvement of 120572PIX in integrin-mediated signalingrdquo HumanMolecular Genetics vol 12 no 2 pp 155ndash167 2003

[89] N A Nejad F Amidi M A Hoseini et al ldquoMale germ-likecell differentiation potential of humanumbilical cordWhartonrsquosjelly-derivedmesenchymal stem cells in co-culture with humanplacenta cells in presence of BMP4 and retinoic acidrdquo IranianJournal of BasicMedical Sciences vol 18 no 4 pp 325ndash333 2015

[90] S S W Chung X Wang S S Roberts S M Griffey P RReczek andD JWolgemuth ldquoOral administration of a retinoicAcid receptor antagonist reversibly inhibits spermatogenesis inmicerdquo Endocrinology vol 152 no 6 pp 2492ndash2502 2011

[91] S Flajollet B Lefebvre C Cudejko B Staels and P LefebvreldquoThe core component of the mammalian SWISNF complexSMARCD3BAF60c is a coactivator for the nuclear retinoic acidreceptorrdquoMolecular and Cellular Endocrinology vol 270 no 1-2 pp 23ndash32 2007

[92] Q H Wu R J Guo M Lin B Zhou and Y F WangldquoMicroRNA-200a inhibits CD1331+ ovarian cancer stem cellsmigration and invasion by targeting E-cadherin repressorZEB2rdquo Gynecologic Oncology vol 122 no 1 pp 149ndash154 2011

[93] F Yu J Li H Chen et al ldquoKruppel-like factor 4 (KLF4) isrequired for maintenance of breast cancer stem cells and for cellmigration and invasionrdquoOncogene vol 30 no 18 pp 2161ndash21722011

[94] K M Bae N N Parker Y Dai J Vieweg and D W SiemannldquoE-cadherin plasticity in prostate cancer stem cell invasionrdquoAmerican Journal of Cancer Research vol 1 pp 71ndash84 2011

[95] A Yasuda H Sawai H Takahashi et al ldquoThe stem cell factorc-kit receptor pathway enhances proliferation and invasion ofpancreatic cancer cellsrdquo Molecular Cancer vol 5 article 462006

[96] M Herreros-Villanueva T K Er and L Bujanda ldquoRetinoicacid reduces stem cell-like features in pancreatic cancer cellsrdquoPancreas vol 44 no 6 pp 918ndash924 2015

[97] A Doi K Ishikawa N Shibata et al ldquoEnhanced expressionof retinoic acid receptor alpha (RARA) induces epithelial-to-mesenchymal transition and disruption of mammary acinarstructuresrdquoMolecular Oncology vol 9 no 2 pp 355ndash364 2015

[98] H-J Yang Y-B Zheng T Ji et al ldquoOverexpression of ILK1 inbreast cancer associates with poor prognosisrdquo Tumor Biologyvol 34 no 6 pp 3933ndash3938 2013

[99] C Leung-Hagesteijn A Mahendra I Naruszewicz and GE Hannigan ldquoModulation of integrin signal transduction byILKAP a protein phosphatase 2C associating with the integrin-linked kinase ILK1rdquoTheEMBO Journal vol 20 no 9 pp 2160ndash2170 2001

[100] J Sanghera P Costello L Kojic et al ldquoIntegrin-linked kinase 1(ILK1) a lsquohotrsquo therapeutic targetrdquo Clinical Cancer Research vol7 pp 3766sndash3767s 2001

[101] D BangWWilsonM Ryan J J Yeh andA S Baldwin ldquoGSK-3120572 promotes oncogenic KRAS function in pancreatic cancervia TAK1-TAB stabilization and regulation of noncanonical NF-120581Brdquo Cancer Discovery vol 3 no 6 pp 690ndash703 2013

[102] P Shi T Yin F Zhou P Cui S Gou and C Wang ldquoValproicacid sensitizes pancreatic cancer cells to natural killer cell-mediated lysis by upregulating MICA and MICB via thePI3KAkt signaling pathwayrdquo BMC Cancer vol 14 article 3702014

[103] J Y C Chow M Ban H L Wu et al ldquoTGF-120573 downregulatesPTEN via activation of NF-120581B in pancreatic cancer cellsrdquoAmer-ican Journal of Physiology Gastrointestinal and Liver Physiologyvol 298 no 2 pp G275ndashG282 2010

[104] L-H Chang S-L Pan C-Y Lai A-C Tsai and C-MTeng ldquoActivated PAR-2 regulates pancreatic cancer pro-gression through ILKHIF-120572-induced TGF-120572 expression andMEKVEGF-A-mediated angiogenesisrdquo The American Journalof Pathology vol 183 no 2 pp 566ndash575 2013

[105] A H Huang S H Pan W H Chang et al ldquoPARVA promotesmetastasis by modulating ILK signalling pathway in lung


adenocarcinomardquo PLoS ONE vol 10 no 3 Article ID e01185302015

[106] Z P Yan H Z Yin R Wang et al ldquoOverexpression ofintegrin-linked kinase (ILK) promotes migration and invasionof colorectal cancer cells by inducing epithelial-mesenchymaltransition viaNF-120581B signalingrdquo Acta Histochemica vol 116 no3 pp 527ndash533 2014

[107] A Traister M Walsh S Aafaqi et al ldquoMutation in integrin-linked kinase (ILK119877211119860) and heat-shock protein 70 comprisea broadly cardioprotective complexrdquo PLoS ONE vol 8 no 11Article ID e77331 2013

[108] J A Gilbert L J Adhikari R V Lloyd T R HalfdanarsonM H Muders and M M Ames ldquoMolecular markers fornovel therapeutic strategies in pancreatic endocrine tumorsrdquoPancreas vol 42 no 3 pp 411ndash421 2013

[109] F Bergmann SAulmann B Sipos et al ldquoAcinar cell carcinomasof the pancreas a molecular analysis in a series of 57 casesrdquoVirchows Archiv vol 465 no 6 pp 661ndash672 2014

[110] LMery B Strauss J FDufour KHKrause andMHoth ldquoThePDZ-interacting domain of TRPC4 controls its localization andsurface expression in HEK293 cellsrdquo Journal of Cell Science vol115 no 17 pp 3497ndash3508 2002

[111] I Del Giudice M Messina S Chiaretti et al ldquoBehind thescenes of non-nodal MCL downmodulation of genes involvedin actin cytoskeleton organization cell projection cell adhe-sion tumour invasion TP53 pathway and mutated status ofimmunoglobulin heavy chain genesrdquo British Journal of Haema-tology vol 156 no 5 pp 601ndash611 2012

[112] X L Li X Lu S Parvathaneni et al ldquoIdentification of RECQ1-regulated transcriptome uncovers a role of RECQ1 in regulationof cancer cell migration and invasionrdquo Cell Cycle vol 13 no 15pp 2431ndash2445 2014

[113] M D Pastor A Nogal S Molina-Pinelo et al ldquoIdentification ofproteomic signatures associated with lung cancer and COPDrdquoJournal of Proteomics vol 89 pp 227ndash237 2013

[114] J D Figueroa S S Han M Garcia-Closas et al ldquoGenome-wide interaction study of smoking and bladder cancer riskrdquoCarcinogenesis vol 35 pp 1737ndash1744 2014

[115] J H Zhou Y J Feng K T Tao et al ldquoThe expression andphosphorylation of ezrin and merlin in human pancreaticcancerrdquo International Journal of Oncology vol 45 no 6 pp2059ndash2067 2014

[116] Y X Meng Z H Lu S N Yu Q A Zhang Y H Ma and JChen ldquoEzrin promotes invasion and metastasis of pancreaticcancer cellsrdquo Journal of Translational Medicine vol 8 article 612010

[117] H M Kocher J Sandle T A Mirza N F Li and I R HartldquoEzrin interacts with cortactin to form podosomal rosettes inpancreatic cancer cellsrdquo Gut vol 58 no 2 pp 271ndash284 2009

[118] Y Oda S Aishima K Morimatsu et al ldquoDifferential ezrin andphosphorylated ezrin expression profiles between pancreaticintraepithelial neoplasia intraductal papillary mucinous neo-plasm and invasive ductal carcinoma of the pancreasrdquo HumanPathology vol 44 no 8 pp 1487ndash1498 2013

[119] J V Thevathasan E Tan H Zheng et al ldquoThe small GTPaseHRas shapes local PI3K signals through positive feedbackand regulates persistent membrane extension in migratingfibroblastsrdquo Molecular Biology of the Cell vol 24 no 14 pp2228ndash2237 2013

[120] H Fensterer K Giehl M Buchholz et al ldquoExpression profil-ing of the influence of RAS mutants on the TGFB1-induced

phenotype of the pancreatic cancer cell line PANC-1rdquo GenesChromosomes amp Cancer vol 39 no 3 pp 224ndash235 2004

[121] F Schonleben W Qiu J D Allendorf J A Chabot H ERemotti and G H Su ldquoMolecular analysis of PIK3CA BRAFand RAS oncogenes in periampullary and ampullary adenomasand carcinomasrdquo Journal of Gastrointestinal Surgery vol 13 no8 pp 1510ndash1516 2009

[122] J Sayyah A Bartakova N Nogal L A Quilliam D G Stupackand J H Brown ldquoThe Ras-related protein Rap1A medi-ates thrombin-stimulated integrin-dependent glioblastoma cellproliferation and tumor growthrdquo The Journal of BiologicalChemistry vol 289 no 25 pp 17689ndash17698 2014

[123] J-S Park K-M Lim S G Park et al ldquoPancreatic cancerinduced by in vivo electroporation-enhanced sleeping beautytransposon gene delivery system in mouserdquo Pancreas vol 43no 4 pp 614ndash618 2014

[124] T Shichinohe N Senmaru K Furuuchi et al ldquoSuppression ofpancreatic cancer by the dominant negative ras mutant N116YrdquoThe Journal of Surgical Research vol 66 no 2 pp 125ndash130 1996

[125] V Sirivatanauksorn Y Sirivatanauksorn and N R LemoineldquoMolecular pattern of ductal pancreatic cancerrdquo LangenbeckrsquosArchives of Surgery vol 383 no 2 pp 105ndash115 1998

[126] A Nagy L Kozma I Kiss et al ldquoCopy number of cancer genespredict tumor grade and survival of pancreatic cancer patientsrdquoAnticancer Research vol 21 no 2 pp 1321ndash1326 2001

[127] T Seufferlein J Van Lint S Liptay G Adler and RM SchmidldquoTransforming growth factor 120572 activates Ha-Ras in human pan-creatic cancer cells with Ki-ras mutationsrdquo Gastroenterologyvol 116 no 6 pp 1441ndash1452 1999

[128] T J Haggerty I S Dunn L B Rose E E Newton FPandolfi and J T Kurnick ldquoHeat shock protein-90 inhibitorsenhance antigen expression on melanomas and increase T cellrecognition of tumor cellsrdquo PLoS ONE vol 9 no 12 Article IDe114506 2014

[129] Y J Zhou and R J Binder ldquoThe heat shock protein-CD91 path-way mediates tumor immunosurveillancerdquo OncoImmunologyvol 3 no 4 Article ID e28222 2014

[130] K Dodd S Nance M Quezada et al ldquoTumor-derivedinducible heat-shock protein 70 (HSP70) is an essential compo-nent of anti-tumor immunityrdquoOncogene vol 34 pp 1312ndash13222015

[131] N Qiao Y Zhu H Li Z Qu and Z Xiao ldquoExpression of heatshock protein 20 inversely correlated with tumor progression inpatients with ovarian cancerrdquo European Journal of Gynaecologi-cal Oncology vol 35 no 5 pp 576ndash579 2014

[132] U Ramp C Mahotka S Heikaus et al ldquoExpression of heatshock protein 70 in renal cell carcinoma and its relation totumor progression and prognosisrdquo Histology and Histopathol-ogy vol 22 no 10ndash12 pp 1099ndash1107 2007

[133] J J Hyun H S Lee B Keum et al ldquoExpression of heat shockprotein 70 modulates the chemoresponsiveness of pancreaticcancerrdquo Gut and Liver vol 7 no 6 pp 739ndash746 2013

[134] Y Kuramitsu Y Wang K Taba et al ldquoHeat-shock protein 27plays the key role in gemcitabine-resistance of pancreatic cancercellsrdquo Anticancer Research vol 32 no 6 pp 2295ndash2299 2012

[135] Z Zhang K Kawamura Y Jiang et al ldquoHeat-shock protein90 inhibitors synergistically enhancemelanoma differentiation-associated gene-7-mediated cell killing of human pancreaticcarcinomardquo Cancer Gene Therapy vol 20 no 12 pp 663ndash6702013


[136] T N MacKenzie N Mujumdar S Banerjee et al ldquoTriptolideinduces the expression of miR-142-3p a negative regulator ofheat shock protein 70 and pancreatic cancer cell proliferationrdquoMolecular Cancer Therapeutics vol 12 no 7 pp 1266ndash12752013

[137] Y K Yu A Hamza T Zhang et al ldquoWithaferin A targetsheat shock protein 90 in pancreatic cancer cellsrdquo BiochemicalPharmacology vol 79 no 4 pp 542ndash551 2010

[138] S Mori-Iwamoto K Taba Y Kuramitsu et al ldquoInterferon-120574down-regulates heat shock protein 27 of pancreatic cancer cellsand helps in the cytotoxic effect of gemcitabinerdquo Pancreas vol38 no 2 pp 224ndash226 2009

[139] S A Lang C Moser A Gaumann et al ldquoTargeting heatshock protein 90 in pancreatic cancer impairs insulin-likegrowth factor-I receptor signaling disrupts an interleukin-6signal-transducer and activator of transcription 3hypoxia-inducible factor-1120572 autocrine loop and reduces orthotopictumor growthrdquo Clinical Cancer Research vol 13 no 21 pp6459ndash6468 2007

[140] VDudeja P Phillips R Sharif Y Zhan RDawra andA SalujaldquoHeat shock protein 70 (HSP70) confers resistance to apoptostsin pancreatic cancer cells by lysosomal stabilizationrdquo Pancreasvol 33 no 4 p 457 2006

[141] W Fendler I Klich A Cieslik-Heinrich K Wyka A Szad-kowska and W Młynarski ldquoIncreased risk of type 1 diabetes inPolish childrenmdashassociation with INS-IGF2 51015840VNTR and lackof associationwithHLAhaplotyperdquoEndokrynologia Polska vol62 no 5 pp 436ndash442 2011

[142] G Sutherland G Mellick J Newman et al ldquoHaplotype anal-ysis of the IGF2-INS-TH gene cluster in Parkinsonrsquos diseaserdquoAmerican Journal of Medical Genetics Part B NeuropsychiatricGenetics vol 147 no 4 pp 495ndash499 2008

[143] C Polychronakos A Kukuvitis N Giannoukakis and EColle ldquoParental imprinting effect at the IPIS-IGF2 diabetessusceptibility locusrdquo Diabetologia vol 38 no 6 pp 715ndash7191995

[144] G Y F Ho A Melman S-M Liu et al ldquoPolymorphism of theinsulin gene is associated with increased prostate cancer riskrdquoBritish Journal of Cancer vol 88 no 2 pp 263ndash269 2003

[145] I A Eaves S T Bennett P Forster et al ldquoTransmission ratiodistortion at the INS-IGF2 VNTRrdquoNature Genetics vol 22 no4 pp 324ndash325 1999

[146] C A Haiman Y Han Y Feng et al ldquoGenome-wide testing ofputative functional exonic variants in rlationship with breastand prostate cancer risk in a multiethnic populationrdquo PLoSGenetics vol 9 no 3 Article ID e1003419 2013

[147] N Kanatsuna A Delli C Andersson et al ldquoDoubly reac-tive INS-IGF2 autoantibodies in children with newly diag-nosed autoimmune (type 1) diabetesrdquo Scandinavian Journal ofImmunology vol 82 pp 361ndash369 2015

[148] N Kanatsuna J Taneera F Vaziri-Sani et al ldquoAutoimmunityagainst INS-IGF2 protein expressed in human pancreatic isletsrdquoJournal of Biological Chemistry vol 288 no 40 pp 29013ndash29023 2013

[149] G Paroni M Fratelli G Gardini et al ldquoSynergistic antitumoractivity of lapatinib and retinoids on a novel subtype of breastcancer with coamplification of ERBB2 and RARArdquo Oncogenevol 31 no 29 pp 3431ndash3443 2012

[150] A Saumet G Vetter M Bouttier et al ldquoTranscriptional repres-sion of microRNA genes by PML-RARA increases expressionof key cancer proteins in acute promyelocytic leukemiardquo Bloodvol 113 no 2 pp 412ndash421 2009

[151] T A Bailey H Luan E Tom et al ldquoA kinase inhibitorscreen reveals protein kinase C-dependent endocytic recyclingof ErbB2 in breast cancer cellsrdquo The Journal of BiologicalChemistry vol 289 no 44 pp 30443ndash30458 2014

[152] H Qi B Sun X Zhao et al ldquoWnt5a promotes vasculogenicmimicry and epithelial-mesenchymal transition via proteinkinase C120572 in epithelial ovarian cancerrdquo Oncology Reports vol32 no 2 pp 771ndash779 2014

[153] A Doller C Winkler I Azrilian et al ldquoHigh-constitutive HuRphosphorylation at Ser 318 by PKC delta propagates tumorrelevant functions in colon carcinoma cellsrdquoCarcinogenesis vol32 no 5 pp 676ndash685 2011

[154] A M Butler M L S Scotti Buzhardt S Li K E Smith AP Fields and N R Murray ldquoProtein kinase C Zeta regulateshuman pancreatic cancer cell transformed growth and invasionthrough a STAT3-dependentmechanismrdquoPLoSONE vol 8 no8 Article ID e72061 2013

[155] Y Chen G Z Yu D H Yu and M H Zhu ldquoPKC120572-induceddrug resistance in pancreatic cancer cells is associated withtransforming growth factor-1205731rdquo Journal of Experimental ampClinical Cancer Research vol 29 no 1 article 104 2010

[156] C E Antal A M Hudson E Kang et al ldquoCancer-associatedprotein kinase C mutations reveal kinasersquos role as tumor sup-pressorrdquo Cell vol 160 no 3 pp 489ndash502 2015

[157] M LinchM Sanz-Garcia E Soriano et al ldquoA cancer-associatedmutation in atypical protein kinase C120580 occurs in a substrate-specific recruitment motifrdquo Science Signaling vol 6 no 293article ra82 2013

[158] P Storz ldquoTargeting protein kinase C subtypes in pancreaticcancerrdquo Expert Review of Anticancer Therapy vol 15 no 4 pp433ndash438 2015

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


tumor progression and invasion [11] Another pair of tumorsuppressor genes SMADDPC4 is involved in PAC throughregulating the TGF-120573 pathway and is critical to advancedtumors [12]

Further heredofamilial pancreatic adenocarcinoma isassociated with certain other significant genes with morecomplexmechanisms As indicated by the available literaturemost such genes participate in the DNA repair process suchas MSH12 PMS12 and BRCA12 which may participatein nonspecific tumor induction [13ndash15] However in severalinherited familial conditions specific known mutations havenot been identified which may hint at the complexity ofcarcinogenicmechanisms and the potential oncogenes aswellas tumor suppressor genes [16]

To predict more potential tumor-related genes we pro-posed a new method that considers protein interactionsfrom STRING (Search Tool for the Retrieval of InteractingGenesProteins) [17] and mines potential PC-related genesSTRING is a database with massive amounts of informationon physical and functional associations between differentproteinsWith an established score system STRINGenables auser to search and browse the protein interactions data as wellas simultaneously quantify the statistic cooccurrence in thebackground [17] Here depending on the method and usinga comprehensive analysis of the protein interaction networkwe fully utilized the database containing reported PC-relatedgenes and predicted potential genes involved in PC

2 Materials and Methods

21 Materials PC-related genes were retrieved from KEGGPATHWAY which is one of the main databases in KEGG(Kyoto Encyclopedia of Genes and Genomes) [18] By exam-ining the pathway hsa05212 pancreatic cancer (httpwwwgenomejpdbget-binwww bgethsa05212 accessed in De-cember 2014) we obtained 65 PC-related genes whichcomprise the gene set S Detailed information on these genesis listed in Supplementary Material I available online athttpdxdoiorg1011552015623121

22 Method for Mining New Candidate Genes in a Protein-Protein Interaction Network Protein-protein interaction(PPI) information is useful for investigating protein-relatedand gene-related problems [19ndash25] Most methods forpredicting protein attributes are based on the notion thattwo proteins that interact always share similar attributes[19ndash23] Because PC-related genes must share commonfeatures related to PC it is reasonable to use PPI informationto identify whether a gene is related to PC Here we adoptedthe PPI information reported in STRING (version 91httpwwwstring-dborg) [17] which is a large onlinedatabase that reports known and predicted protein interac-tions The PPIs reported in STRING are derived from thefollowing four sources (1) genomic context (2) high-throughput experiments (3) (conserved) coexpression and(4) previous knowledge which imply that the physical andfunctional associations of the proteins were measured Toextract PPI information for humans we downloaded the file

ldquoproteinlinksv91txtgzrdquo and selected protein interactionsbeginning with ldquo9606rdquo which produced 2425314 humanPPIs Each PPI includes two proteins and one score whichmeasures the strength of the interaction in the range between150 and 999 For the formulation let us denote the score ofan interaction between proteins 119901

1and 119901

2as 119876119894(1199011 1199012)

The aforementioned PPIs were used to construct a largenetwork by taking proteins as nodes Two nodes are adjacentif and only if the corresponding proteins comprise an inter-action that is a member of the 2425314 PPIs Furthermorethe interaction score should also be added to the constructednetwork To generate compatibility between our network anda shortest path approach each edge e was assigned a weight119908(119890) as follows 119908(119890) = 1000 minus 119876

119894(1199011 1199012) where 119901

1and 119901

2

are corresponding proteins of the endpoints of 119890Two proteins that can interact with each other always

share similar attributes [19ndash23] Further considering theinteraction score this notion can be generalized as followstwo proteins in an interaction with a high score are morelikely to share similar attributes than those with a low scoreMoreover if we consider a series of proteins 119901

1 1199012 119901

119899

such that two consecutive proteins can comprise an inter-action with a high score then these proteins may all sharesome common attributes By mapping these proteins ontothe constructed network they may comprise a shortest pathconnecting 119901

1and 119901

119899 which is a path connecting 119901

1and 119901

119899

such that the summation of the weights of edges on the pathis minimumThus we searched all shortest paths connectingany two PC-related genes For two consecutive nodes in eachof these paths their corresponding proteins can comprise aPPI with a high score because they lie on a shortest pathAs mentioned above they can share similar functions Fora specific shortest path 119901

1 1199012 119901

119899 where 119901

1and 119901

119899are

encoded byPC-related genes1199012shares similar functionswith

1199011 that is 119901

2may be encoded by an invalidated or known

PC-related gene 1199013shares similar functions with 119901

2 thus 119901

3

may also share similar functions with 1199011andmay be encoded

by an invalidated or known PC-related gene This can beinduced to 119901

4 1199015 119901

119899minus1 Thus we extracted genes in these

shortest paths and excluded those that were members of 65PC-related genes These genes were referred to as shortestpath genes and deemed to have special relationships with PCThe similar scheme has been applied to extract novel genesor chemicals related to other diseases or some biologicalprocesses [26ndash29] In addition to distinguish those geneseach shortest path genewas assigned a value which is referredto as betweenness and defined as the number of shortest pathsthat contain the gene The betweenness indicates the directand indirect associations between shortest path genes andPC-related genes [30]

23 Further Selection By executing themethodmentioned inSection 22 certain shortest path genes can be extracted fromthe constructed network However certain such genes maybe false positives To exclude this type of gene the followingmethod was adopted

Certain nodes in the constructed network are generalhubs their corresponding genes may always receive a high




1198782 119878




119894 and the











3

6

1

1

3

1

2

1

1

1

11

1

1

1

11

1

1

1

1

3

1

10

2

6

1

1

1

11

5

2

1

2

1

1

1

1

1

3

1

1

1

1

1

1

1

1

1

1

1

21

1

11

1

1

2

1

2

1

2

1

1

1

1

13

6

8

1

1

1

2

1

1

1

1

11

1

1

2

1

1

1

6

5

1

1

1

1

11

33

1

1

4

1

2

1

1

1

1

1

11

2

5

1

1

1

3

2

1

1

1 1 14

2

2

1

2

1

1

1

2

1

1

1

1

1

4

2

2

3

3

1

28

6

1

1

1

1

1

1

2

1 13

1

1

11

1

1

1

1

11

1

11 1

1

1

3

1 11

1

3

1

11

21

11 1

2

1

1

1

1 1

1

2

1

1

1

1

11

11

111

1

1

1 1 1

1

1 1 1

1

1

1

1

1

1

3

1

1

1

2

2

1

11

1

1

1 1

1

111

1

35

1

1

1

1

11

1

1

1

1

2

2 11

1 1

1

1

1 11 11

1

11

2

1

1

1

11

ENSP00000282561

ENSP00000361120

ENSP00000295400ENSP00000366563

ENSP00000384675

ENSP00000278568

ENSP00000287647

ENSP00000293288ENSP00000366244

ENSP00000340944

ENSP00000019317

ENSP00000355896

ENSP00000361125

ENSP00000005257

ENSP00000375892

ENSP00000330237

ENSP00000329623

ENSP00000269300ENSP00000352121ENSP00000321410

ENSP00000268182ENSP00000278616

ENSP00000302564ENSP00000369497

ENSP00000362649

ENSP00000355249ENSP00000269571ENSP00000345571

ENSP00000266970ENSP00000361423

ENSP00000265171

ENSP00000267101

ENSP00000267868

ENSP00000417281

ENSP00000284384

ENSP00000262904ENSP00000262741

ENSP00000288986

ENSP00000320940

ENSP00000262435ENSP00000329357

ENSP00000332973

ENSP00000298316

ENSP00000347858

ENSP00000371067

ENSP00000358022

ENSP00000003084

ENSP00000250894

ENSP00000269321ENSP00000288602

ENSP00000355153

ENSP00000274335

ENSP00000384515

ENSP00000338018

ENSP00000348461

ENSP00000338934

ENSP00000343204ENSP00000263025

ENSP00000304895

ENSP00000303830ENSP00000348986

ENSP00000289153

ENSP00000297494

ENSP00000309103

ENSP00000304283

ENSP00000262160

ENSP00000254066ENSP00000354394

ENSP00000309845

ENSP00000268058

ENSP00000341551

ENSP00000244007

ENSP00000046794

ENSP00000302269

ENSP00000269305

ENSP00000215832

ENSP00000302486ENSP00000226574

ENSP00000251849ENSP00000264033

ENSP00000353483ENSP00000228872ENSP00000352157

ENSP00000350941

ENSP00000162330

ENSP00000228307

ENSP00000360683ENSP00000206249

ENSP00000339151ENSP00000358622

ENSP00000262367

ENSP00000270202ENSP00000257904ENSP00000221930

ENSP00000359424ENSP00000384273

ENSP00000263826ENSP00000364133

ENSP00000222254ENSP00000314458

ENSP00000263967ENSP00000351905

ENSP00000360266ENSP00000309503

ENSP00000354558ENSP00000300161ENSP00000219476

ENSP00000335153ENSP00000216797ENSP00000350283

ENSP00000263253

ENSP00000272519

ENSP00000262613

ENSP00000227507

ENSP00000342793ENSP00000344818

ENSP00000349467

ENSP00000265734

ENSP00000222005ENSP00000261799

ENSP00000267163ENSP00000238682

ENSP00000264657

ENSP00000223023

ENSP00000249071

ENSP00000341189

ENSP00000299421

ENSP00000244741

ENSP00000339007

ENSP00000250617

ENSP00000344456

ENSP00000359206

ENSP00000275493ENSP00000300574












128 0031
















4 Conclusions




Acknowledgments


References













































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




1198782 119878




119894 and the











3

6

1

1

3

1

2

1

1

1

11

1

1

1

11

1

1

1

1

3

1

10

2

6

1

1

1

11

5

2

1

2

1

1

1

1

1

3

1

1

1

1

1

1

1

1

1

1

1

21

1

11

1

1

2

1

2

1

2

1

1

1

1

13

6

8

1

1

1

2

1

1

1

1

11

1

1

2

1

1

1

6

5

1

1

1

1

11

33

1

1

4

1

2

1

1

1

1

1

11

2

5

1

1

1

3

2

1

1

1 1 14

2

2

1

2

1

1

1

2

1

1

1

1

1

4

2

2

3

3

1

28

6

1

1

1

1

1

1

2

1 13

1

1

11

1

1

1

1

11

1

11 1

1

1

3

1 11

1

3

1

11

21

11 1

2

1

1

1

1 1

1

2

1

1

1

1

11

11

111

1

1

1 1 1

1

1 1 1

1

1

1

1

1

1

3

1

1

1

2

2

1

11

1

1

1 1

1

111

1

35

1

1

1

1

11

1

1

1

1

2

2 11

1 1

1

1

1 11 11

1

11

2

1

1

1

11

ENSP00000282561

ENSP00000361120

ENSP00000295400ENSP00000366563

ENSP00000384675

ENSP00000278568

ENSP00000287647

ENSP00000293288ENSP00000366244

ENSP00000340944

ENSP00000019317

ENSP00000355896

ENSP00000361125

ENSP00000005257

ENSP00000375892

ENSP00000330237

ENSP00000329623

ENSP00000269300ENSP00000352121ENSP00000321410

ENSP00000268182ENSP00000278616

ENSP00000302564ENSP00000369497

ENSP00000362649

ENSP00000355249ENSP00000269571ENSP00000345571

ENSP00000266970ENSP00000361423

ENSP00000265171

ENSP00000267101

ENSP00000267868

ENSP00000417281

ENSP00000284384

ENSP00000262904ENSP00000262741

ENSP00000288986

ENSP00000320940

ENSP00000262435ENSP00000329357

ENSP00000332973

ENSP00000298316

ENSP00000347858

ENSP00000371067

ENSP00000358022

ENSP00000003084

ENSP00000250894

ENSP00000269321ENSP00000288602

ENSP00000355153

ENSP00000274335

ENSP00000384515

ENSP00000338018

ENSP00000348461

ENSP00000338934

ENSP00000343204ENSP00000263025

ENSP00000304895

ENSP00000303830ENSP00000348986

ENSP00000289153

ENSP00000297494

ENSP00000309103

ENSP00000304283

ENSP00000262160

ENSP00000254066ENSP00000354394

ENSP00000309845

ENSP00000268058

ENSP00000341551

ENSP00000244007

ENSP00000046794

ENSP00000302269

ENSP00000269305

ENSP00000215832

ENSP00000302486ENSP00000226574

ENSP00000251849ENSP00000264033

ENSP00000353483ENSP00000228872ENSP00000352157

ENSP00000350941

ENSP00000162330

ENSP00000228307

ENSP00000360683ENSP00000206249

ENSP00000339151ENSP00000358622

ENSP00000262367

ENSP00000270202ENSP00000257904ENSP00000221930

ENSP00000359424ENSP00000384273

ENSP00000263826ENSP00000364133

ENSP00000222254ENSP00000314458

ENSP00000263967ENSP00000351905

ENSP00000360266ENSP00000309503

ENSP00000354558ENSP00000300161ENSP00000219476

ENSP00000335153ENSP00000216797ENSP00000350283

ENSP00000263253

ENSP00000272519

ENSP00000262613

ENSP00000227507

ENSP00000342793ENSP00000344818

ENSP00000349467

ENSP00000265734

ENSP00000222005ENSP00000261799

ENSP00000267163ENSP00000238682

ENSP00000264657

ENSP00000223023

ENSP00000249071

ENSP00000341189

ENSP00000299421

ENSP00000244741

ENSP00000339007

ENSP00000250617

ENSP00000344456

ENSP00000359206

ENSP00000275493ENSP00000300574












128 0031
















4 Conclusions




Acknowledgments


References













































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


3

6

1

1

3

1

2

1

1

1

11

1

1

1

11

1

1

1

1

3

1

10

2

6

1

1

1

11

5

2

1

2

1

1

1

1

1

3

1

1

1

1

1

1

1

1

1

1

1

21

1

11

1

1

2

1

2

1

2

1

1

1

1

13

6

8

1

1

1

2

1

1

1

1

11

1

1

2

1

1

1

6

5

1

1

1

1

11

33

1

1

4

1

2

1

1

1

1

1

11

2

5

1

1

1

3

2

1

1

1 1 14

2

2

1

2

1

1

1

2

1

1

1

1

1

4

2

2

3

3

1

28

6

1

1

1

1

1

1

2

1 13

1

1

11

1

1

1

1

11

1

11 1

1

1

3

1 11

1

3

1

11

21

11 1

2

1

1

1

1 1

1

2

1

1

1

1

11

11

111

1

1

1 1 1

1

1 1 1

1

1

1

1

1

1

3

1

1

1

2

2

1

11

1

1

1 1

1

111

1

35

1

1

1

1

11

1

1

1

1

2

2 11

1 1

1

1

1 11 11

1

11

2

1

1

1

11

ENSP00000282561

ENSP00000361120

ENSP00000295400ENSP00000366563

ENSP00000384675

ENSP00000278568

ENSP00000287647

ENSP00000293288ENSP00000366244

ENSP00000340944

ENSP00000019317

ENSP00000355896

ENSP00000361125

ENSP00000005257

ENSP00000375892

ENSP00000330237

ENSP00000329623

ENSP00000269300ENSP00000352121ENSP00000321410

ENSP00000268182ENSP00000278616

ENSP00000302564ENSP00000369497

ENSP00000362649

ENSP00000355249ENSP00000269571ENSP00000345571

ENSP00000266970ENSP00000361423

ENSP00000265171

ENSP00000267101

ENSP00000267868

ENSP00000417281

ENSP00000284384

ENSP00000262904ENSP00000262741

ENSP00000288986

ENSP00000320940

ENSP00000262435ENSP00000329357

ENSP00000332973

ENSP00000298316

ENSP00000347858

ENSP00000371067

ENSP00000358022

ENSP00000003084

ENSP00000250894

ENSP00000269321ENSP00000288602

ENSP00000355153

ENSP00000274335

ENSP00000384515

ENSP00000338018

ENSP00000348461

ENSP00000338934

ENSP00000343204ENSP00000263025

ENSP00000304895

ENSP00000303830ENSP00000348986

ENSP00000289153

ENSP00000297494

ENSP00000309103

ENSP00000304283

ENSP00000262160

ENSP00000254066ENSP00000354394

ENSP00000309845

ENSP00000268058

ENSP00000341551

ENSP00000244007

ENSP00000046794

ENSP00000302269

ENSP00000269305

ENSP00000215832

ENSP00000302486ENSP00000226574

ENSP00000251849ENSP00000264033

ENSP00000353483ENSP00000228872ENSP00000352157

ENSP00000350941

ENSP00000162330

ENSP00000228307

ENSP00000360683ENSP00000206249

ENSP00000339151ENSP00000358622

ENSP00000262367

ENSP00000270202ENSP00000257904ENSP00000221930

ENSP00000359424ENSP00000384273

ENSP00000263826ENSP00000364133

ENSP00000222254ENSP00000314458

ENSP00000263967ENSP00000351905

ENSP00000360266ENSP00000309503

ENSP00000354558ENSP00000300161ENSP00000219476

ENSP00000335153ENSP00000216797ENSP00000350283

ENSP00000263253

ENSP00000272519

ENSP00000262613

ENSP00000227507

ENSP00000342793ENSP00000344818

ENSP00000349467

ENSP00000265734

ENSP00000222005ENSP00000261799

ENSP00000267163ENSP00000238682

ENSP00000264657

ENSP00000223023

ENSP00000249071

ENSP00000341189

ENSP00000299421

ENSP00000244741

ENSP00000339007

ENSP00000250617

ENSP00000344456

ENSP00000359206

ENSP00000275493ENSP00000300574












128 0031
















4 Conclusions




Acknowledgments


References













































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology









128 0031
















4 Conclusions




Acknowledgments


References













































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology








4 Conclusions




Acknowledgments


References













































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


































































































































































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

research article mining for candidate genes...

Documents